alinaqi/claude-bootstrap
12 skills10.9k installs8.3k starsGitHub
Install
npx skills add https://github.com/alinaqi/claude-bootstrapSkills in this repo
1Android KotlinAndroid Kotlin is a path-triggered agent skill for solo builders shipping native Android apps with Jetpack Compose, Hilt dependency injection, Kotlin coroutines, and MockK-backed unit tests. It does not replace your IDE—it steers the coding agent when it opens Kotlin sources or Gradle Kotlin scripts so changes follow a conventional clean-architecture folder layout, feature-scoped screens and ViewModels, and separated data versus domain boundaries. Use it throughout Build when you are implementing screens, wiring repositories, or adjusting module Gradle files. Medium effort is expected because Android touches manifest, resources, DI graphs, and tests together. Ideal for one-person shops standardizing on Google's recommended stack without pasting stack-overflow fragments into every chat.2.4kinstalls2Pwa Developmentpwa-development is a frontend-focused agent skill for solo and indie builders who want reliable, fast, installable web experiences without shipping a native app store binary first. It frames Progressive Web Apps around three non-negotiable requirements—HTTPS (with localhost for dev), a background service worker for offline and cache control, and a web app manifest that describes how the app installs and appears on devices. The SKILL.md orients you toward practical implementation concerns: which assets to cache, how offline fallbacks should behave, and how Workbox fits into a maintainable caching strategy rather than ad-hoc fetch interceptors. Use it when you are touching service worker files, manifest metadata, or Workbox configuration and need procedural guardrails so your agent does not skip security or installability basics. It pairs naturally with a shipping phase performance pass but belongs on the build shelf because the artifacts are authored in the frontend codebase.1.9kinstalls3Playwright TestingPlaywright Testing is a procedural E2E skill for solo builders shipping web apps who need fast, cross-browser confidence without reinventing config every sprint. It documents installation for greenfield and brownfield repos, a production-style Playwright config with CI-aware retries, workers, and reporters (HTML, list, GitHub), and baseURL-driven runs against localhost or staged URLs. The guidance emphasizes reliability patterns from Playwright best practices—traces on first retry, screenshots and video retained on failure, and a setup project for authentication that runs once ahead of dependent suites. Page Object structuring and path-scoped invocation (`e2e`, spec files, playwright folder) keep agents focused when you are fixing flaky selectors or standing up a new suite. Invoke it whenever you are authoring specs or debugging failures in the browser layer before launch.1.8kinstalls4Ui MobileUi-mobile is a React Native-focused agent skill from the Claude Bootstrap kit that turns mobile accessibility from optional advice into enforceable patterns. Solo builders shipping iOS and Android apps invoke it while editing TSX screens so every tappable control respects a 44-point minimum size and text meets WCAG 2.1 AA contrast on light and dark surfaces. The skill calls out common failures—undersized hit areas, gray-on-gray labels—and supplies copy-paste StyleSheet structures for buttons and icon buttons. It applies when you are building or refactoring mobile UI, not when you are choosing product scope or writing store listings. Pair it with your component library work so agents do not shrink controls for aesthetics and leave real users with failed taps and illegible secondary text.857installs5Posthog AnalyticsPosthog-analytics is an agent skill for solo and indie builders who need PostHog wired correctly from day one. It walks through implementing product analytics with a clear rule: measure what matters—activation, retention, where users drop off, which features get used, and whether acquisition is working—rather than logging every click. The skill covers installation for Next.js App Router, initializing the client with environment keys, choosing identified-only profiles, and handling pageviews manually for single-page apps. It also points you at feature flags and dashboards so experiments and KPI views stay aligned with the same event model. Use it when you are adding analytics, feature flags, or event tracking with PostHog during build or right after launch, and you want agent guidance that matches PostHog’s own documentation patterns instead of ad-hoc copy-paste snippets.655installs6FlutterFlutter is an agent skill aimed at solo and indie builders shipping cross-platform mobile apps who want agents to respect a consistent Dart codebase instead of improvising folder chaos. It encodes a layered structure—core utilities, data and domain repositories, and presentation features with dedicated providers and widgets—centered on Riverpod for state, Freezed for immutable models, go_router for navigation, and mocktail-friendly tests. The skill is path-triggered whenever you touch Dart sources, pubspec.yaml, or tests, so it acts as a standing style guide during active Build work rather than a one-off generator. That makes it especially useful when you iterate features quickly with Claude Code or Cursor and need routing, theming, and provider patterns to stay uniform across modules. It does not replace platform-specific release checklists; it keeps implementation discipline inside lib/ while you focus on product screens and API wiring.615installs7Code ReviewADR Gate is a low-effort Ship-phase checker for solo and indie builders who use structured /code-review workflows in Claude Code and similar agents. Before any review engine analyzes your diff, it classifies whether the change is trivial enough to skip architectural linkage or requires ADR and spec context. Non-trivial work triggers discovery of existing decision records and specifications, injection into review context, or presentation of a reverse-engineered ADR draft tagged as proposed. Trivial edits—typos, markdown-only updates, test-only changes that do not alter behavior, and non-structural config tweaks—short-circuit the gate so you are not blocked on ceremony. The skill encodes a explicit decision tree and optional shell patterns so agents behave consistently instead of debating whether a change is architectural. Use it whenever you want reviews grounded in why the system is shaped the way it is, not just whether the patch compiles.614installs8Supabase Nextjssupabase-nextjs is a build-phase skill for solo developers shipping SaaS on Next.js with Supabase as the managed backend and Drizzle as the type-safe query layer. It encodes the App Router project skeleton: auth routes under `(auth)`, dashboard segments, API route handlers, shared UI components, and a `src/db` area for schema and query functions. The guiding rule keeps Drizzle responsible for SQL-shaped data access while Supabase clients handle authentication, storage, and realtime, with server components preferred and client components only where interactivity demands it. Invoke it when `when-to-use` matches—building a Next.js app with a Supabase backend across `src/app`, `src/db`, and `supabase` paths. It reduces architectural drift for one-person teams who need SSR-friendly auth without mixing ORM and Supabase responsibilities. Expect medium effort: migrations, env keys, and RLS policies still require your attention beyond the template.570installs9SupabaseSupabase is a workflow-oriented agent skill that teaches the standard CLI loop indie builders should use for Postgres-backed SaaS and agent products: init locally, link a remote project, express schema and RLS in migrations, and promote through pipelines instead of dashboard surgery. It maps when to lean on Auth, Storage, Realtime, Edge Functions, and pgvector so you do not bolt on redundant services. Solo developers invoke it while editing supabase folders or environment files so the agent suggests commands and patterns aligned with official Supabase docs. The skill emphasizes medium effort changes—migrations, policies, and function scaffolding—not a full greenfield tutorial. Pair it with your CI job definitions so local supabase db push or deploy steps match what runs in GitHub Actions or similar.496installs10Android Javaandroid-java is a path-triggered Claude Bootstrap skill for solo developers shipping native Android apps in Java. It encodes a conventional clean-ish layout: data and remote layers, domain use cases, Hilt or Dagger DI, and MVVM feature folders with Activities, Fragments, and ViewModels, plus res layouts and manifest placement. ViewBinding and Espresso testing are emphasized so agents do not mix deprecated patterns or scatter UI logic across layers. The skill auto-applies when you touch Java sources or Gradle under android paths, which keeps recommendations consistent across refactors and new features. It is intermediate in scope—assumes Android Studio, Gradle, and standard app module structure—and targets the Build phase when you are actively coding screens, repositories, and instrumentation tests rather than validating a product idea.484installs11React NativeReact Native is a build-phase agent skill that steers solo and indie builders toward maintainable mobile app structure when an agent touches React Native sources. It activates on TypeScript and JavaScript UI files plus iOS, Android, and app.json paths, signaling medium-effort guidance rather than one-off snippets. The documented layout keeps pure business logic under src/core, reusable UI under components with tests, and route-level screens with colocated hooks and barrel indexes. Navigation, shared hooks, state, and utils get explicit homes so agents do not scatter platform code. Component patterns emphasize simple functional components with typed props—Pressable and Text examples—instead of legacy class patterns. For a solo builder shipping a cross-platform app with Claude Code, Cursor, or Codex, the skill reduces inconsistent folder sprawl and makes incremental features easier to review. It does not replace Expo or native toolchain setup; it complements day-to-day implementation once the project exists.455installs12Code DeduplicationCode Deduplication is an agent skill for solo and indie builders who want lean codebases while moving fast with Claude Code, Cursor, or similar agents. It installs a check-before-you-write habit backed by a CODE_INDEX.md capability map so the agent knows what already exists before inventing another formatter, validator, or HTTP wrapper. The philosophy is explicit: AI tends to reimplement intent, so the guardrail is against duplicate purpose, not just identical text. Before any new function or shared utility, the workflow is search the index, search the codebase, prefer extension, and only then add net-new code—then refresh the index in the same session. A periodic /audit-duplicates pass catches drift when multiple contributors or long threads pile on helpers. It pairs well with planning and review skills but does not replace tests or formal refactors. Use it whenever the repo is growing helper surface area; skip it for throwaway spikes you will delete the same day.1installs